Lymphocytic myocarditis may be idiopathic (also called primary lymphocytic or autoreactive myocarditis) or may be secondary to viral infection (for example, enterovirus Coxsackie B infection), other infections, connective tissue disease, sarcoidosis, and medications [2, 3]. On the basis of the results of animal experiments, it is thought that most cases of idiopathic lymphocytic myocarditis are caused by a virus, even though that virus has not been identified. In animal models, the inflammatory response incited by the virus is directed against the virus, then develops against the myocytes themselves. This autoimmune process persists even after the virus can no longer be identified. [2]. Immunosuppressive agents have long been considered in the treatment of myocarditis to minimize inflammatory cell-mediated damage. We focus on the effectiveness of immunosuppressive treatments in lymphocytic myocarditis.

The role of immunosuppressive therapy in lymphocytic myocarditis is controversial. Seven factors contribute to the difficult task of undertaking and interpreting clinical research in this field.

1. The incidence of recognized symptomatic myocarditis is unknown but is probably low [4], limiting human studies.

2. Myocarditis varies in presentation, ranging from acute fulminant congestive heart failure to chronic idiopathic dilated cardiomyopathy. As a consequence, it is difficult to generalize treatment response to the whole spectrum of the disease.

3. Immunosuppressive agents may influence inflammatory mediators differently and have various treatment results [2]. For example, we refer to all treatments as immunosuppressive therapies in this article, but the effects of prednisone and cyclosporine are very different than those of immunomodulating treatments, such as specific immunoglobulins and thymic thyromodulin.

4. Numerous viruses are implicated in myocarditis, and treatment response may differ. For example, patients with adenovirus-positive myocarditis may benefit more from immunoglobulin treatment than do patients with myocarditis stemming from other viral causes [5].

5. Myocarditis affects persons of all ages, and disease and treatment response may vary between children and adults.

6. Previous randomized trials seem to show that myocarditis improves with conservative treatment that includes other proven therapies for congestive heart failure [6]. As a result, it is difficult to interpret the effectiveness of immunosuppressive therapy in studies that did not include a control group.

7. History, physical examination, and routine investigations have poor specificity for myocarditis. In 1984, standardized microscopic criteria were developed for the interpretation of endomyocardial biopsies to aid in diagnosis and to facilitate research (Dallas criteria). This historical gold standard serves as the basis for diagnosis in this article, but questions about its validity are discussed [7].

Our goal was to consolidate the best existing evidence from human clinical trials of immunosuppressive therapy in lymphocytic myocarditis. Taking into account the confounding variables previously identified, we also address the limitations of our present knowledge.

Methods

Top Methods Data Analysis Discussion Author & Article Info References

Locating Research

Two of the authors independently searched the MEDLINE, CINAHL, HEALTHSTAR, and Cochrane Collaboration databases from 1980 to 26 June 1997 (standardized criteria for the assessment of endomyocardial biopsy were not determined until 1986 [5]). Paired medical subject headings used for keyword and text word searching were immunosuppression, immunotherapy, azathioprine, cyclosporine, interferon, globulin, cyclophosphamide, antivirals, levamisole, thymic hormone, corticosteroid, or prednisone in myocarditis or cardiomyopathies. We also searched for articles that included any drug treatment in myocarditis. Reference lists of articles that fulfilled our inclusion criteria were scanned to identify articles not found in searches of the computerized databases. A cardiologist with an interest in myocarditis was consulted to ensure that relevant studies were not missed. Reliability between authors was good but was not quantified.

Selecting Research

The following inclusion criteria were used to select studies for the overview.

Design

Randomized clinical trials, case–control studies, or cohort studies that compared one or more immunosuppressive regimens with each other or with no treatment were included. All patients could continue receiving conventional therapy (digoxin, diuretic agents, or vasodilators) in addition to immunosuppressive agents. Neither the patients nor the physicians caring for the patients were required to be blinded to treatment assignments. The reports included in the analysis were limited to English-language articles. Single-cohort studies were excluded because of methodologic concerns and because the natural history of left ventricular ejection fraction in myocarditis is one of improvement [6], a fact that confounds positive responses attributed to immunotherapy in single cohorts.

Target Population

Patients of all ages with biopsy-proven endomyocardial myocarditis were included. Biopsy results of acceptable evidence of myocarditis with conventional staining were 1) Dallas criteria of definitive or borderline myocarditis or 2) a mean of 2.5 lymphocytes or more per high-power field (x400) [7]. Patients were excluded if they had medical conditions that might mimic myocarditis or dilated cardiomyopathy (such as coronary artery disease, the postpregnancy state, and ethanol abuse).

Outcome

The study had two selected outcome measures: mortality, as defined by death or the need for cardiac transplantation, and change in resting left ventricular ejection fraction, as assessed by echocardiography, cardiac catheterization, or radionuclide ventriculography. Outcomes not considered in this analysis include resolution of infiltrate on repeated biopsy and exercise tolerance testing.

Two of the authors reviewed the titles and abstracts of all printouts and independently applied the three criteria described above to the citations. All articles selected by one or both of the authors were retrieved, and the selection criteria were independently reapplied to the full text of the articles to make a final selection.

Assessment of Methodologic Quality

We independently reviewed the trials and analyzed their validity by using methodologic criteria [8-10]. Five criteria were applied that are known to be potential sources of bias: trial randomization, blinding, concealment of randomization, matched co-interventions, and proper follow-up. When opinions on the same report differed, the report was discussed among the authors and a final determination was made.

Data Analysis

Top Methods Data Analysis Discussion Author & Article Info References

Changes in left ventricular ejection fraction were reported as the mean change with an associated 95% CI. When the CI was not provided in the original report, the CI of the difference between means was calculated by using the SEs of before and after measurements. It should be noted that the SE is larger than expected because the correlation between pre- and post-test measurements was not provided. The Fisher exact test was used to determine whether the mortality rate differed between groups.

Data Synthesis

Study Location, Study Selection, and Assessment of Methodologic Quality

From among 374 references, 12 relevant randomized and case–control trials were identified. Our inclusion criteria represented subgroups of 3 reports; therefore, only results applying to these subgroups were analyzed [11-13]. One report was excluded because only two patients in the study fulfilled our inclusion criteria [14]. One report was excluded because the trial in question is ongoing, but this study is mentioned in the discussion section [15] and is identified in Table 1 and Table 2. Three reports [13, 18, 19] were accounts of the same study, and 2 reports [6, 20] both reported on another study. One retrospective case–control report on the use of -globulin was excluded because the data limited to biopsy-proven myocarditis could not be extracted [21]. The resulting 6 trials, along with the assessment of methodologic quality, are summarized in Table 1 and Table 2 [6, 11-13, 15-17].

Control Group

One randomized, controlled trial [6] demonstrated an improvement in mean left ventricular function at 28 weeks (change in left ventricular ejection fraction, 0.07 [95% CI, 0.03 to 0.12]) in the control group. Similarly, three matched cohort trials [12, 13, 17] showed improvements in left ventricular function at 2 to 89 months (change in left ventricular ejection fraction, 0.10 [CI, –0.03 to 0.23]), 24 weeks (change, 0.04 [CI, 0.01 to 0.09]), and 35 weeks (change, 0.02 [CI, –0.06 to 0.10]). Only one small matched-cohort trial [16] showed decreased left ventricular function in the control group at 3 months (change in left ventricular ejection fraction, –0.03[CI, –0.08 to 0.02]).

Immunosuppressive Agents

Prednisone

A small, randomized, controlled trial did not show a decrease in mortality with prednisone between groups [11]. The prednisone arm from a small pediatric matched-cohort trial demonstrated no change (P > 0.2) in left ventricular ejection fraction between the prednisone group (change, 0.05 [CI, –0.03 to 0.13]) and the control group (change, 0.02 [CI, –0.06 to 0.10]) [17].

Prednisone and Azathioprine or Prednisone and Cyclosporine

The Myocarditis Treatment Trial [6], a multicenter, randomized, controlled trial in adults, did not show any difference in improvement in left ventricular ejection fraction between patients receiving treatment with prednisone and azathioprine or prednisone and cyclosporine (change, 0.10 [CI, 0.07 to 0.12]) compared with the control group (change, 0.07 [CI, 0.03 to 0.12]) (P > 0.2). Similarly, mortality at 1 year did not differ significantly between the treatment group (in which 15 of 47 patients died) and the control group (in which 15 of 64 patients died) (P > 0.2). These results are similar to those of a small, single-center, retrospective, matched-cohort trial [12] that showed no significant change in left ventricular ejection fraction between a group that received prednisone alone and a group that received prednisone and azathioprine (change, 0.09 [CI, –0.03 to 0.21]; P = 0.11) and a control group (change, 0.10 [CI, –0.02 to 0.23]; P = 0.10). Once again, mortality did not differ significantly between the treatment groups (in which 2 of 8 patients died) and the control group (in which 5 of 10 patients died) (P > 0.2). A single-center trial from Germany [16] reported a significant change in left ventricular ejection fraction at 3 months in 38 adults receiving conventional therapy (change, –0.03[CI, –0.07 to 0.02]) or prednisone and azathioprine (change, 0.08 [CI, –0.09 to 0.24]). However, when one considers the large CI, the calculated change in left ventricular ejection fraction was not significant (P > 0.2).

The results of the above trials differ from those of a small, single-center, pediatric matched-cohort trial [17] that demonstrated a significant improvement in left ventricular ejection fraction in patients who received prednisone and azathioprine (change, 0.28 [CI, 0.22 to 0.34]; P < 0.01) and those who received prednisone and cyclophosphamide (change, 0.32 [CI, 0.24 to 0.34]; P < 0.01) compared with a control group (change, 0.02 [CI, –0.06 to 0.10]; P > 0.2) (P < 0.01).

Thymus Extract or Interferon-

A small, single-center, matched-cohort trial [13] showed no significant difference in left ventricular ejection fraction in patients who received interferon (change, 0.07 [CI, 0.02 to 0.12]) or thymus thyromodulin (change, 0.04 [CI, –0.01 to 0.90]) compared with a control group (change, 0.08 [CI, 0.03 to 0.13]). No deaths occurred in any of the groups.

Discussion

Top Methods Data Analysis Discussion Author & Article Info References

Myocarditis is a clinically important disease for which the use of immunosuppressive therapy varies widely. This overview summarizes the literature to facilitate clinical decision making and improve consensus on the use of this therapy. The best evidence to date suggests that immunosuppressive therapy is ineffective for myocarditis. Prednisone alone is not effective in adult or pediatric myocarditis. Similarly, prednisone combined with azathioprine or cyclosporine is not effective in adult myocarditis. Our experience with other therapies, such as interferon or thymus extract, are limited, but initial results are not promising.

One matched-cohort trial differed from all other studies by demonstrating a benefit of prednisone combined with azathioprine or cyclosporine in children [17]. This discordance may be due to methodologic weaknesses. First, although the trial claimed to be randomized, patients were assigned to treatment groups on the basis of hospital admission, making the study a nonconcealed matched-cohort trial. Second, although no patients seemed to have been lost to follow-up, 7 patients were not properly accounted for and were not analyzed with the groups to which they were assigned; 43 patients were enrolled in the trial, but 50 were accounted for in the treatment groups. Given the size of the study, these patients could have altered the results from a significant to an insignificant difference between groups. Finally, with respect to matched co-interventions, it is not clear whether the groups were similar in all aspects other than hemodynamic status; this could have also altered results. With these limitations, it is difficult to support the use of immunosuppressive treatment in children until further research is available.

If immunosuppressive therapy is not effective for myocarditis, what treatment options remain? Myocarditis varies in its presentation, ranging from chronic idiopathic dilated cardiomyopathy to acute fulminant congestive heart failure. In the studies discussed above, patients who have symptoms of myocarditis for 4 weeks to 2 years seemed to improve with conservative treatment alone. Most studies used a conventional regimen of reduced salt intake, digoxin, diuretic agents, angiotensin-converting enzyme inhibitors, nitrates, hydralazine, warfarin, or antiarrhythmic agents. With conventional therapy, the left ventricular ejection fraction seemed to improve approximately 10% after 6 months. Unfortunately, patients with acute fulminant myocarditis were not included in the trials, but the data give no reason to believe that the natural progression of myocarditis in survivors of fulminant myocarditis would differ. As a result, this group should be given aggressive supportive therapy, including intensive care admission, inotropic agents, intra-aortic balloon pumps, or extracorporeal membrane oxygenation. For all presentations of myocarditis, cardiac transplantation is a treatment option and should be considered in end-stage heart disease.

Given that treatment options for myocarditis are limited to conservative therapies that already represent standard treatment for congestive heart failure, the question is whether endomyocardial biopsy is necessary in the management of these patients. Endomyocardial biopsy of the right ventricle performed with a bioptome introduced into the internal jugular vein has associated risks. With no effective specific treatment available for lymphocytic myocarditis, a strong argument can be made to treat conservatively and not pursue the diagnosis. Furthermore, concerns have been raised about the validity of endomyocardial biopsy in identifying patients with the disease. From a technical standpoint, biopsy of the right ventricle may miss focal areas of disease. In addition, diagnosis has traditionally been made by a nonspecific light microscopic pattern; for lymphocytic myocarditis, this pattern is a lymphocytic infiltrate with possible myocyte necrosis. Therefore, it is not surprising that low interobserver reliability has been described for the Dallas criteria [6]. For these reasons, it is a less-than-ideal gold standard. In addition, no evidence suggests that the amount of inflammation correlates with outcome. Thus, for diagnostic, treatment, and prognostic reasons, endomyocardial biopsy seems to have no role in the management of lymphocytic myocarditis. Experience with other, rarer forms of myocarditis, such as giant-cell myocarditis, is limited, and no randomized studies have been done. However, immunosuppressive therapy seems to be similarly ineffective in these cases [1]. As a result, the present use of endomyocardial biopsy in myocarditis is limited to research purposes.

It is hoped that future research will elucidate effective treatment for myocarditis. To facilitate this, diagnostic criteria that incorporate etiology may help identify populations that may respond better to specific therapy. New tools, such as immunohistochemistry, polymerase chain reaction, in situ hybridization, and Southern blot analysis of blood and tissue, may assist in the identification of active viral disease that may respond to antiviral agents. Similarly, the absence of virus may better identify an autoimmune form of myocarditis in which present and future immunosuppressive and immunomodulating agents may be most effective. New information on myocarditis is forthcoming with the results of the European Study of Epidemiology and Treatment of Cardiac Inflammatory Disease trial (ESETCID) [15], an ongoing, multicenter, randomized, placebo-controlled, double-blind study (Table 1 and Table 2) that uses some of the diagnostic tools discussed above to divide patients with biopsy-proven myocarditis into three subgroups-those with autoimmune, enterovirus, and cytomegalovirus myocarditis. According to communications with the chief investigator of ESETCID, 50 patients were enrolled as of April 1997, and results may be available as early as 1999. Finally, future research in myocarditis may be facilitated through the development of international registries, which may better document and describe clinical experiences with this uncommon disease.

Conclusions

Immunosuppressive therapy is ineffective in lymphocytic myocarditis in both adults and children. Current therapy seems to be limited to supportive measures or transplantation. With conservative therapy, the natural course of the disease is one of improvement. It is hoped that future research will elucidate more effective treatment for myocarditis by identifying its causes more clearly, making it possible to tailor therapy appropriately.